Biology subjects

Singh, R.

Publications and source records attributed to Singh, R..

At least 19 recordsLinked to original sources

bioRxiv2026

Computational Prediction of Recombinant Expression Efficiency for Industrial Enzymes: An Integrated Co-Translational Sequence Analysis Platform

Industrial biocatalysis relies on the cost effective production of active enzymes in heterologous hosts, yet a substantial fraction of candidate sequences fails during recombinant expression because sequence-level determinants of translational efficiency are not captured by conventional codon optimization alone. Here we present an integrated computational platform for pre-clinical assessment of industrial enzyme expression potential that combines baseline sequence metrics with co-translational modeling of ribosome stalling, translation rhythm, codon harmonization, folding bottlenecks, solubility proxies, and proteolysis risk. The platform accepts DNA, protein, FASTA, and PDB inputs and returns interpretable expression risk assessments, component-level diagnostics, variant design proposals, and actionable engineering recommendations. Analyses are parameterized for industrial expression hosts (Escherichia coli, Saccharomyces cerevisiae, Bacillus subtilis, and mammalian cells) and for native enzyme sources including Streptomyces, Bacillus, filamentous fungi, actinomycetes, and extremophiles. All analyses execute locally without dependence on external databases, enabling rapid and secure screening of proprietary enzyme candidates prior to gene synthesis and fermenter scale-up. We report three layers of preliminary validation on small cohorts: (i) a retrospective public industrial-enzyme ranking benchmark (n = 20; rank subset n = 10); (ii) calibration of the solubility module on held-out eSOL enzymes with quantitative experimental solubility; and (iii) an independent collaborator wet-lab construct set retained as a true external-test holdout and never used for model fitting (n = 38 labeled constructs). On the external holdout, solubility calls achieved accuracy 68.4% (95% CI 52.5 - 80.9%), sensitivity 56.5% (36.8 - 74.4%), specificity 86.7% (62.1 - 96.3%), balanced accuracy 71.6% (bootstrap 95% CI 57.4 - 84.4%), and MCC 0.43 (0.15 - 0.69). Sensitivity remains close to chance on this small set, whereas specificity is higher; intervals are wide and the evidence should be read as directional rather than definitive. A directional arginine-surface redesign example (SrMilN insoluble parent to SrMilN_SASAR partially soluble variant; incomplete rescue with multiple soluble bands) was recovered. These findings support preliminary screening utility and transparent experimental agreement, but do not constitute high-accuracy prediction of industrial fermenter titre. Offline EC-class heuristics are retained only as optional provisional labels and are not presented as a validated platform capability (agreement 3/20 = 15%, 95% CI 5.2 - 36%; chance for six classes is 16.7%).

bioinformatics↗

bioRxiv2026

The p53-p21-Cyclin D2 regulatory axis drives metabolic reprogramming and a distinct senescent macrophage senotype during aging and MASLD.

Aging drives chronic disease in part through senescent cells, including macrophages, which fuel inflammation. Senescent macrophages are functionally heterogeneous: canonical p16-high macrophages promote tumorigenesis or, in other contexts, disease tolerance, whereas we previously identified a distinct p21-high, p16-low senotype that drives metabolic dysfunction-associated steatotic liver disease (MASLD). The molecular basis of this senotype has remained undefined. Using genetic and multi-omic approaches, we show that a p53-p21-dependent program actively represses p16 and is required for senescent macrophage viability. We identify Cyclin D2 as a non-canonical downstream effector that redistributes from the nucleus to mitochondria and lipid droplets, where it partners with MIC60 to drive metabolic reprogramming and modulates AKT1-mTORC1 signaling that sustains the SASP. Cyclin D2-p21 double-positive macrophages accumulate with aging and MASLD in mice and in human liver cirrhosis, and can be selectively depleted by senolytic treatment. Together, these findings define a druggable p53-p21-Cyclin D2 axis that specifies macrophage senotype.

immunology↗

bioRxiv2026

TFE3 fusions drive expression of CD44 and SPP1 in Translocation Renal Cell Carcinoma

Xp11.2 translocation RCC [Xp11.2 tRCC]) is an underdiagnosed and aggressive subtype of RCC with few specific or targeted therapies. The transmembrane glycoprotein CD44 is an emerging target in many advanced malignancies, and with its ligand OPN (SPP1), is a crucial driver of cancer progression, stemness, metastasis, and immune suppression. Here we show that common TFE3-fusions [ including SFPQ-TFE3, PRCC-TFE3, ASPSCR1-TFE3, and NONO-TFE3] are associated with upregulated expression of CD44 and SPP1, as observed in multiple human and murine bulk transcriptomic studies and a murine tRCC snRNA-Seq dataset. CD44 and/or SPP1 protein expression were also upregulated in murine models of transgenic tRCC kidneys and urine specimens, patient-derived cell lines and human tRCC cases, by immunoblotting and/or IHC. Transient deletion of CD44 was associated with profound and specific suppression of tRCC cell line growth, with decreased mTOR signaling. These data suggest that CD44 and/or SPP1 may potentially drive tumorigenesis in tRCC.

cancer biology↗

bioRxiv2026

A Hybrid Residual-Swin Transformer Design with Attention for Prostate Cancer Segmentation

Prostate cancer, a leading cause of cancer-related deaths among men globally, necessitates the development of precise diagnostic and treatment strategies. Accurate segmentation of prostate cancer in medical imaging, particularly in MRI scans, is crucial for early diagnosis and clinical decisions. Traditional manual segmentation techniques, while efficient, are labor-intensive and necessitate significant expertise, resulting in an increasing demand for automated alternatives. The RSAUNet is a novel deep learning architecture designed to improve prostate cancer segmentation. It incorporates essential components, including Residual Blocks, Swin Transformer Blocks, and Attention Mechanisms within a U-Net architecture. This markedly enhances the model's capacity to discern complex anatomical features and accurately segment malignant tissues. Using sophisticated deep Learning methods, RSAUNet addresses the complexities of prostate imaging, delivering reliable, consistent segmentation results. The model was evaluated against various cutting-edge techniques on extensive multi-parametric MRI datasets, attaining an impressive Dice coefficient (DC) of 0.998 and a Jaccard Index (IoU) of 0.965. These findings highlight the innovative characteristics of RSAUNet and its capacity to transform prostate cancer diagnosis and treatment strategies. The proposed model surpasses current methods and shows potential for practical clinical applications, providing an efficient, precise, and scalable solution for automated prostate cancer segmentation and fostering optimism for the future of medical imaging and diagnosis.

cancer biology↗

bioRxiv2026

A Simple Method to Distinguish Active and Inactive Aptamers by Analyzing the Ruggedness of the Aptamer Free Energy Landscape

Aptamers are structured nucleic acid ligands capable of high affinity, high specificity molecular recognition generated using variations of the SELEX (Systematic Evolution of Ligands by Exponential Enrichment) process. However, SELEX often produces sequences that enrich yet may lack binding efficacy. We propose a measure called the Ruggedness Composite Index (RCI) along with a method for computing it, that can be used to distinguish binding-competent ('active') aptamers from weak or non-binding ('inactive') aptamers. Given a set of aptamers, RCI incorporates information on their fragmentation (landscape partitioning), basin entropy (metastable state distribution), cumulative density irregularity (non-uniform occupancy), and structural energy correlation length (structure-energy coupling scale). We test whether secondary-structure folding energy landscape topology distinguishes active from inactive aptamers using a multiscale level set framework across six datasets. Active aptamers show lower RCI values and occupy smoother, funnel-like conformational spaces, while inactive aptamers show higher RCI values, reflecting fragmented, high-entropy landscapes. By contrast, classical thermodynamic features, such as minimum free energy, show limited discrimination between active and inactive aptamers. In all datasets, sequences that exhibit enrichment which is not monotonic but lack specificity exhibit elevated ruggedness, indicating landscape topology can predict non-specific enrichment. These results indicate that folding landscape organization can be used as a predictor of aptamer activity and establish RCI as a simple, mechanistically interpretable measure for improving candidate prioritization, especially in therapeutic aptamer discovery.

bioinformatics↗

bioRxiv2026

Design and Application of a μSiM Outer Blood-Retinal Barrier (OBRB) Model as a Drug Development Tool

The outer blood-retinal barrier (OBRB) is the primary interface through which systemically circulating drugs reach the retina. A tool that measures delivery across this barrier would support the development of targeted therapies as alternatives to repeated intravitreal injection, and the screening of drugs that reach the retina as an off-target toxicity. Such a tool should deliver drugs fluidically through a vascular compartment, measure transport across the retinal pigment epithelium (RPE), and display disease phenotypes relevant to efficacy. Here we adapt the SiM platform, which places epithelium and endothelium in direct juxtaposition across a permeable, optically transparent silicon nitride nanomembrane. ARPE-19 and human umbilical vein endothelial cells (HUVECs) were used as development cell sources. ARPE-19 monocultures reached a transepithelial electrical resistance of 68 {+/-} 26 {Omega} cm2 by 28 days, and ARPE-19 + HUVEC co-cultures reached a small-molecule permeability of 6.34 {+/-} 1.3 x 10-4 cm min-1 within 14 days, a state reported elsewhere only after longer culture. The barriers developed an intervening basement membrane. Drugs perfused through the basal vascular channel crossed into an open apical well, where sampling and mass spectrometry showed transport correlating with lipophilicity, as reported in vivo. The device also displayed two clinically relevant phenotypes. Digoxin at a clinically toxic concentration reduced viability in the co-barrier by about half and doubled permeability. In a vascularized configuration, VEGF drove endothelial invasion of the RPE layer, as seen in neovascular AMD. The SiM-OBRB therefore satisfies basic design criteria for measurement of drug bioavailability, toxicity, and efficacy.

bioengineering↗

bioRxiv2026

A novel ciprofloxacin analogue enables daptomycin-mediated killing of resistant Staphylococcus aureus by increasing septum formation

Quinolone antibiotics such as ciprofloxacin inhibit DNA gyrase, leading to DNA double-strand breaks that result in rapid bacterial killing and induction of the mutagenic SOS DNA repair response. By contrast, the ciprofloxacin analogue IMP-1700 inhibits ciprofloxacin-induced SOS, suggesting a novel mechanism of action. Here, we provide evidence that IMP-1700 targets the quinolone binding domain of DNA gyrase but triggers a significantly higher frequency of division septa in S. aureus compared with other DNA gyrase targeting antibiotics, including ciprofloxacin. In keeping with this finding, the lipopeptide antibiotic daptomycin, which targets the division septum, bound more strongly to IMP-1700-treated cells relative to S. aureus exposed to other DNA gyrase inhibitors, leading to increased bacterial killing. This finding extended to a panel of paired daptomycin susceptible and resistant clinical isolates. We conclude that the ciprofloxacin analogue IMP-1700 has distinct effects on the cell envelope of S. aureus, despite appearing to share the same target as the parent drug, which result in the resensitisation of daptomycin resistant bacteria to the lipopeptide antibiotic.

microbiology↗

bioRxiv2026

Plasma Metabolomic Profiling of COPD Patients Stratified by Smoking Status: A GC-MS- Based Approach

BackgroundChronic obstructive pulmonary disease, primarily caused by exposure to cigarette smoke, is a heterogeneous lung condition characterized by complex metabolic alterations. The metabolic changes associated with smoking status have not been thoroughly investigated. Our study aims to explore the metabolite profile of COPD patients categorised by their smoking habits, including smokers, ex-smokers, and non-smokers. MethodsIn this study, the plasma metabolome of smoking stratified COPD patients were assessed using gas chromatography coupled to mass spectrometry. We applied multivariate and univariate statistical analysis to identify the differentially abundant metabolites. ResultsWe identified 23 altered metabolites in the smokers and 36 in the ex-smokers COPD subgroups. Interestingly, in comparison to the control group, no significant alteration was observed in the plasma of non-smoker COPD patients. Additionally, pathway enrichment analysis revealed top dysregulated metabolic pathways, including biosynthesis of unsaturated fatty acids, galactose metabolism, phenylalanine, tyrosine, and tryptophan biosynthesis, and glycosylphosphatidylinositol (GPI)-anchor biosynthesis. The receiver operating characteristic curve screened five metabolites, such as tetradecanoic acid, 2,4-di-tert-butylphenol, chloroxylenol, tetradecanal, and 1-dodecene, with the highest diagnostic performance (AUC > 0.8). ConclusionThis study reveals distinct plasma metabolic signatures across COPD subgroups categorized by cigarette smoking history.

biochemistry↗

bioRxiv2026

Sequence-dependent conformational and mechanical landscapes of double-stranded nucleic acids

The sequence-dependent mechanical landscapes of double-stranded nucleic acid (dsNA) remain largely unexplored beyond canonical dsDNA. We describe cgNA+, a coarse-grained predictive model of the mechanics of dsRNA, DNA:RNA hybrids, and epigenetically modified dsDNA, all parameterised from 1.26 milliseconds of atomistic simulations. cgNA+ predicts non-local sequence-dependent equilibrium shape and stiffness with errors an order of magnitude smaller than sequence-variability, while enabling exploration of numbers of sequences inaccessible to atomistic simulation. We show that dsNA equilibrium shape is strongly influenced by flanking sequence up to octamer context, with flexible dimer-steps more context-sensitive. CpG-modification alters equilibrium shape comparable to changes caused by single-nucleotide polymorphisms. Groove width analysis across dsNA decamers reveals strong sequence dependence, reflecting the differing characteristic helical geometry of dsDNA and dsRNA, whereas DRHs exhibit mixed behaviour depending on DNA-strand pyrimidine content. CTCF binding sites exhibit a distinct groove width signature. Persistence-length spectra from [~] 9 million sequences indicate that dsRNA is stiffer than dsDNA, whereas DRH exhibit intermediate stiffness modulated by DNA strand pyrimidine content. Persistence length increases upon CpG-modification, but decreases on hypermodification. Overall, the cgNA+ model enables a first, highly accurate, very large-scale, comparative study of sequence-dependent mechanics both within and across dsNA classes, demonstrating previously hidden regulatory layers. Graphical Abstract O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=74 SRC="FIGDIR/small/740023v1_ufig1.gif" ALT="Figure 1"> View larger version (25K): org.highwire.dtl.DTLVardef@1a430edorg.highwire.dtl.DTLVardef@88469borg.highwire.dtl.DTLVardef@d3a7eforg.highwire.dtl.DTLVardef@1b65a15_HPS_FORMAT_FIGEXP M_FIG C_FIG

biophysics↗

bioRxiv2026

COMPASS: Component-Wise Inference of Shared and Gene-Specific Perturbation Response

Predicting how a genetic perturbation reshapes a cells transcriptome is a central goal of computational biology. Previous studies report that the mean response across training perturbations rivals specialized models on standard accuracy metrics, even though it cannot distinguish which perturbation occurred. Across 2,270 CRISPRi perturbations measured in each of six cell lines, we show that this apparent paradox reflects a conserved organization of perturbation responses. Perturbations span a continuum from responses strongly aligned with the mean to more targeted responses that depart from it. Crucially, a perturbations position along this continuum is conserved across cell lines (Kendalls W = 0.59) and predictable from STRING protein-interaction embeddings (R2 = 0.35). We formalize this structure with COMPASS, an interpretable linear model that decomposes each response into shared and gene-specific components and estimates them separately. The shared-response component is modeled as a cell-line-wide response scaled by a perturbation-specific coefficient. This coefficient is strongly conserved across cell lines. The residual gene-specific component--which is moderately conserved across cell lines--recovers pathway-level programs. COMPASS outperforms scGPT, CPA, GEARS, GenePert, and SO_SCPLOWTATEC_SCPLOW in both response accuracy (de-biased Pearson delta 0.34 vs. [≤] 0.32) and perturbation discrimination (cosine PDS gain 0.23 vs. [≤] 0.08). These results recast perturbation prediction across cellular contexts as component-wise inference, with each component estimated from the evidence best suited to it.

bioinformatics↗

bioRxiv2026

Transcriptomic Illuminations: How light intensity reshapes the Chlamydomonas reinhardtii cell cycle

The unicellular algae cell cycle can be divided into several phases, including the commitment point (CP), a point-of-no-return where the cell decides to divide, presumably based on reaching a critical cell size. Light plays a crucial role in the fitness of photosynthetic algal cells, affecting both CP timing and the number of daughter cells produced. So far, only few genes involved in CP have been described, and the presumed sizer and its signal(s) remain unidentified. Using synchronized cells and varying light intensities as a proxy, we explored the effects of light intensity in Chlamydomonas reinhardtii and observed both physiological and transcriptional changes occurring before and after CP under low light (LL) (100 {micro}mol m-{superscript 2} s-{superscript 1}) and optimal light (OL) (500 {micro}mol m-{superscript 2} s-{superscript 1}) conditions. Although CP was delayed by approximately 6 hours in LL, resulting in smaller mother cells and fewer daughter cells, the cells divided at the same time in both conditions. Overall, nucleic acid, protein, and energy reserve levels were lower in LL, with almost no starch produced. RNA-seq analysis identified several core genes shared between both conditions, with 201 genes expressed only in pre-CP1, 161 genes specific to post-CP1, and 582 shared across different phases. In LL, RNA-seq analysis showed an increase in differentially expressed genes (DEGs) in pre-CP1 compared to post-CP1, with an emphasis on photosynthesis, RNA metabolism, and organelle production before commitment, and cell division-related pathways (microtubules, DNA recombination) after CP1. In OL, the number of DEGs increased in post-CP1 by approximately 41% compared to pre-CP1, with a strong emphasis on protein production throughout the cell cycle.

cell biology↗

bioRxiv2026

Integrated solid/solution NMR assignment allows mapping dynamics and ligand binding in a 134 kDa enzyme

Understanding enzyme function requires characterizing not only static structure but also dynamics and ligand interactions. NMR spectroscopy provides this insight at atomic resolution, yet for large proteins the difficulty of resonance assignment has largely confined such studies to systems below[~] 50 kDa, or to observing only methyl groups. Here we present an integrated magic-angle spinning (MAS) and solution NMR study of the 134 kDa tetrameric malate dehydrogenase from Ignicoccus islandicus (IiMDH), an enzyme of particular interest as an evolutionary intermediate between allosteric lactate dehydrogenases and non-allosteric malate dehydrogenases. By combining high-dimensional (up to 4D) MAS NMR experiments on sedimented protein with solution NMR, we achieved 92% backbone heavy-atom assignment and 91% assignment of all Ile-{delta}1, Leu-{delta}1/-{delta}2, Val-{gamma}1/-{gamma}2, Met-{varepsilon} and Thr-{gamma} methyl groups. Building on these assignments, MAS NMR 15N rotating-frame relaxation (R1p) measurements revealed pronounced microsecond-timescale backbone dynamics in functionally critical regions, including the catalytic loop and the mobile surface loop. Complementary methyl-axis order parameters from solution NMR identified additional flexible sites in the hydrophobic core. Chemical shift perturbation experiments upon addition of the substrate analogue oxamate, monitored via backbone 1H-15N TROSY, revealed both active-site contacts and rearrangements of helices a2F and a3G, regions implicated in allosteric signal transmission. The integrated approach demonstrated here exploits the distinct strengths of MAS and solution NMR, and provides a comprehensive view of structure, dynamics, and substrate interactions in a large oligomeric enzyme that would not be accessible by either technique alone.

biophysics↗

bioRxiv2026

scWeave: A deep learning model that bidirectionally translates between gene expression and chromatin structure at single cell resolution

Chromatin structure and gene expression are intimately linked, yet characterizing how the two covary has proven challenging, primarily because the two modalities are rarely measured in the same cells. Recently, single-cell co-assay protocols have enabled simultaneous profiling of both modalities within the same cells, but these experiments remain costly and technically challenging. To better characterize the relationship between 3D chromatin architecture and gene expression and to enable cross-modality inference from single-modality measurements, we developed a model called scWeave that bidirectionally translates between gene expression (scRNA-seq) and 3D chromatin architecture (scHi-C) at single-cell resolution. The scWeave model employs dual autoencoders to extract separate cell-level latent representations and learns to translate between these representations using dedicated translation modules. We evaluate scWeave on six publicly available co-assay datasets spanning mouse embryonic development, mouse cortex, mouse olfactory epithelium, and human bone marrow. On held-out mouse cells, scWeave outperforms a nearest-neighbor baseline and existing methods adapted to single-cell resolution, achieving a 57% improvement in median Spearman correlation when predicting gene expression from chromatin structure and an 18.8% improvement in median HiCRep similarity in the reverse direction relative to the next-best baseline. We further show that scWeave learns cross-modally aligned latent representations at single-cell resolution, enabling cells profiled in one modality to be matched to their counterparts in the other. Finally, scWeave generalizes to entirely held-out developmental timepoints in mouse olfactory epithelium and performs well on held-out human bone marrow cells despite limited human training data. By predicting the unmeasured chromatin architecture or transcriptional state from a single measured modality, scWeave offers a route to extend the benefits of costly co-assays to the many cell types, developmental stages, and species that are currently profiled with only one modality.

genomics↗

bioRxiv2026

TP53-mediated bidirectional lineage plasticity drives alveolar epithelial cell extrusion and tissue remodeling

Cell extrusion contributes to epithelial homeostasis, but its dysregulation can lead to tumorigenesis or degeneration. A fine balance in this process is therefore essential for tissue integrity. Yet the cell types and states vulnerable to extrusion, and the mechanisms that drive it, remain elusive. Here, using spatial maps of cell states in human idiopathic pulmonary fibrosis (IPF) we find that aberrant TP53 activation in alveolar epithelial cells drives cell extrusion. Genetic modulation of TP53 specifically in alveolar epithelial type 1 cells (AT1) was sufficient to induce plasticity and subsequent extrusion as demonstrated by lineage tracing and live imaging. Strikingly, single cell and bulk transcriptome profiling revealed aberrant TP53 drives AT1 cells to acquire a transitional state mirroring AT2-derived regeneration associated intermediate states. Critically, loss of AT1 derived transitional state triggers a compensatory AT2-derived regenerative response, establishing a bidirectional transitional state that activates myofibroblasts and remodels the alveolus. Together, our study implicates AT1 plasticity and their reversion as an unrecognized driver of epithelial cell loss and establishes bidirectional transitional state as a central mechanism underlying progression of fibrotic remodeling.

Cell Biology↗

bioRxiv2026

Immunoengineered Chitosanase-Produced Chitosan Oligomers for Elevating Plant Resistance to Viral Infection

Chitooligomers can act as plant biostimulants or biopesticides, but todays chitosan-based agro-biologics often lack sufficient efficacy. This is due to a lack of scalable production processes for structurally well-controlled chitosans combined with a limited understanding of structure-function relationships. Chitosans differ in their degree of polymerization (DP), fraction and pattern of acetylation (FA and PA). While the influence of DP and FA on antimicrobial and phytostimulatory properties is at least partially known, this is not yet the case for PA. PA can be partially controlled by using enzymatic rather than acid hydrolysis for oligomer production. We have used recombinant chitinases and chitosanases to hydrolyse a well-characterised chitosan polymer, and purified oligomers with different DP. We have structurally characterised the products and tested their abilities to protect tobacco from viral disease. Chitinase products were dominated by GlcNAc units at their reducing and non-reducing ends, with GlcN units dominating their centers, and v.v. for chitosanase products. While the chitinase-derived hydrolysates were inactive, the chitosanase-derived oligomers possessed elicitor and priming activities and protected plants from disease, and their activity increased with increasing DP. Clearly, the Bacillus chitosanase used is well-suited to set up a scalable production process for chitosan oligomers with promising agro-biologic properties. Graphical Abstract O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=95 SRC="FIGDIR/small/731087v1_ufig1.gif" ALT="Figure 1"> View larger version (56K): org.highwire.dtl.DTLVardef@b50b59org.highwire.dtl.DTLVardef@1fc35c3org.highwire.dtl.DTLVardef@12db91aorg.highwire.dtl.DTLVardef@d3732f_HPS_FORMAT_FIGEXP M_FIG C_FIG

plant biology↗

bioRxiv2026

Reconstruction of septin higher-order nano-size structures in ovarian cancer cells uncover susceptibility to the septin-targeting small molecule UR214-9

In cancer cells, septins assemble into enigmatic higher-order structures of 300-700 nanometers, including long needle-like filaments, thick perinuclear rings, and cytoplasmic bundles or aggregates. The absence of genetic or pharmacological tools to recapitulate these architectures in-vitro has impeded mechanistic studies of their formation, function, and therapeutic targeting. Here, first, determining the overexpression of septin-2 in epithelial ovarian cancer (EOC) and its association with increased mortalities and dependencies, we select SKOV-3 ovarian cancer cells as a tractable model in which septin supramolecular assemblies can be recreated in-vitro and interrogated. This system shows that the forchlorfenuron (FCF) analog UR214-9 remodels septin architecture, converting co-expressed human septin octamers (SEPT2-SEPT6-SEPT7-SEPT9-SEPT9-SEPT7-SEPT6-SEPT2) into large cytoplasmic aggregates. In parallel, transiently expressed SEPT2 is reorganized into septin-rich noodle-like filaments, perinuclear rings, and web-like networks encircling the nucleus upon UR214-9 treatment. Mechanistically, UR214-9 disrupts the incorporation of SEPT2, SEPT7, and SEPT9 into canonical septin hetero-octamers, resulting in assembly-defective or imperfect oligomers that preferentially reorganize into these aberrant higher-order structures. This aggregation likely prevents septin-2 migration during interphase-to-cleavage furrow transition in NRK-49F-SEPT2-EGFP homozygous cells and impacts SKOV-3 cytokinesis, cell proliferation, adhesion and invasion and migration while sparing ceramide transport to the Golgi, preserving ER and cis-Golgi structure. These effects manifested in reduced growth of ovarian, endometrial and breast cancer xenografts without attracting significant off-target engagements per the global transcriptomic analysis of JIMT1 breast cancer and PANC-1 pancreatic cells. UR214-9 treated animals showed observable safety in animals. Thus, a tool to recreate aberrant septin structures and identification of septins as a druggable cytoskeletal target for ovarian, endometrial, breast and pancreatic cancer by perturbing their hetero-octamerization assembly is presented. SignificanceWe provide a method to reconstruct the higher-order septin architecture observed in cancer cells, to study their assembly and functions. Intriguingly, cancer cells tolerate hetero-oligomeric septins lacking specific subunits, suggesting that compositionally deficient oligomers are not efficiently targeted for degradation, unlike unincorporated septin monomers in normal cells. This tolerance may enable accumulation of structurally aberrant septin complexes acquiring long-needles, rings or thick-aggregates in disease cells. We further show that septin oligomerization can be pharmacologically perturbed. By integrating structural, cellular, and energetic readouts using in-silico techniques, we establish a quantitative framework for septin-targeted modulation, generating UR214-9 as a new chemotype that disrupts septin oligomeric assembly via preventing incorporation of SEPT2/7/9, into canonical hetero-octamers, causes defects in cytokinesis, altered cell migration, viability, and remodels septin-actin architectures, ultimately impairing tumor cell growth. Thus, pharmacological targeting of septin assembly represents a tractable strategy to perturb septin-dependent cellular processes in cancer and neurodegenerative diseases with reported septin dysregulation.

cancer biology↗

bioRxiv2026

Occurrence of Biased mTOR Signaling in Hepatocellular Carcinoma

BackgroundmTOR signaling promotes cell growth and anabolic processes in all eukaryotes. Hyperactivation of mTOR signaling is associated with various cancers along with hepatocellular carcinoma (HCC). HCC is a highly lethal malignancy with multiple aetiologies such as viral infection, alcohol abuse, and metabolic dysfunction. Therapeutic options for HCC remain limited due to an incomplete understanding of oncogenic drivers and poorly characterized mechanisms of disease progression. MethodsVarious regimens of DEN and CCl4 carcinogen dosage were investigated on C57BL/6J mice to induce HCC. The histological analysis for fibrosis and serum markers for liver function were performed. Molecular analyses of oncogenic drivers were performed in the HCC tissues obtained from mice and HCC patients. The impact of inhibition of mTOR signaling was assessed on HCC progression. ResultsWe established a rapid DEN+CCl4 induced (DCI) HCC model in C57BL/6 mice to study disease progression longitudinally. The molecular analysis revealed upregulation of MAPK and downregulation of mTORC1-S6K-S6 signaling in HCC. However, other branches of mTOR such as mTORC1-ULK1, mTORC1-4EBP1, and mTORC2-PKC were upregulated due to the increased expression. Similar observations were found in tissues derived from HCC patients. Furthermore, inhibition of mTORC1 alone by Rapamycin did not reduce HCC progression but Torin 1 mediated inhibition of both mTORC1 and mTORC2 significantly reduced HCC progression. ConclusionsWe propose this biased mTOR signaling modulates mTOR activity towards specific downstream processes that are crucial for cancer cell growth and targeting both the mTOR complexes has better therapeutic potential in HCC. Impact and ImplicationsThis study provides a rapid pre-clinical model for understanding HCC progression and to explore various intervention strategies. The study reports a novel phenomenon of biased mTOR signaling where deregulation of downstream substrate levels modulates the mTOR activity towards the specific branches, the master regulator of cell growth and metabolism. Furthermore, the study suggests that the clinical investigations exploring the rapalogs against HCC should be cautiously considered depending on the aetiology and signaling status of HCC. Graphical abstract O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=125 SRC="FIGDIR/small/727188v2_ufig1.gif" ALT="Figure 1"> View larger version (20K): org.highwire.dtl.DTLVardef@ef3714org.highwire.dtl.DTLVardef@1091b25org.highwire.dtl.DTLVardef@d52f2forg.highwire.dtl.DTLVardef@cc158_HPS_FORMAT_FIGEXP M_FIG C_FIG HighlightsO_LIDEN and CCl4 treatment generated a well-established HCC within 4 months. C_LIO_LILiver fibrosis and serum markers correlated with HCC progression. C_LIO_LIUpregulation of mTOR pathway substrates create biased mTOR signaling. C_LIO_LIDual inhibition of mTORC1 and mTORC2 reduced HCC progression significantly. C_LI

cancer biology↗

bioRxiv2026

A Multimodal Neural Network Model for Early Recurrence Prediction in Lung Adenocarcinoma

Lung adenocarcinoma (LUAD), a subtype of non-small cell lung cancer (NSCLC), is the most common primary lung cancer worldwide. Despite advancements in early detection and treatment, up to 39% of patients develop recurrent tumors following complete resection. Currently, no widely available models exist for reliably predicting early recurrence of LUAD, which is a significant prognostic factor of post-recurrence survival. Models leveraging deep learning (DL) techniques have demonstrated notable utility in cancer recurrence prediction, particularly when used in combination with both clinical and genomic data. We developed a DL-based model, Predicting Lung Adenocarcinoma recurrence via Selective Multimodal Attention (PLASMA), to predict early recurrence using clinical, mRNA expression, and mutation data from patients with primary stage I-III LUAD. Trained on The Cancer Genome Atlas (TCGA) dataset, PLASMA outperformed traditional machine learning models in predicting early recurrence in both the TCGA test set and an external validation set (TRACERx Lung), achieving area under the receiver operating characteristic curve (AUROC) scores of 85.0% and 76.5%, respectively. Our results support the potential of multimodal DL for early LUAD recurrence prediction and risk stratification.

bioinformatics↗